Article Text

Download PDFPDF

Early symptoms of brain tumours
  1. E Davies1,
  2. C Clarke2
  1. 1Department of Palliative Care and Policy, Guy’s, King’s and St Thomas’ School of Medicine, Weston Education Centre, Cutcombe Road, Denmark Hill, London SE5 9RJ, UK
  2. 2National Hospital for Neurology and Neurosurgery, University College Hospitals, Queen Square, London WC1N 3BG, UK
  1. Correspondence to:
 Dr Elizabeth Davies

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Malignant cerebral glioma is the most common adult primary brain tumour but surprisingly few studies report how patients with early symptoms present in primary or secondary care. A retrospective audit in south east Scotland found considerable variation in the referral of patients with primary brain tumours: only one quarter of 439 patients were initially referred directly to specialist centres.1 This must relate in part to the way in which symptoms develop and the difficulty of distinguishing them from more common but less sinister problems. For example, a large case record review of initial symptoms experienced by 653 glioma patients presenting to the National Hospital for Neurology and Neurosurgery, Queen Square, London, between 1955 and 1975 found a relatively low prevalence of neurological problems such as epilepsy (38%), headache (35%), mental change (17%), and hemiparesis (10%); by the time of diagnosis the prevalences were 54%, 71%, 52%, and 43%, respectively.2 Few studies focus on the accounts of patients and relatives. One qualitative interview study of 28 Swedish patients suggested that relatives noticed general changes including cognitive and personality change and took the initiative in seeking help more often than the patients themselves.3

The recently published last diaries of the politician and historian the late Alan Clark provoked us to reconsider the significance of early symptoms from the perspective of patients and their close relatives.4 Clark provides us with a moving account of the gradual onset of symptoms from a glioma—fatigue, problems with thinking and concentration, and intermittent headache over nine months. He also describes vividly the anxiety of knowing something was wrong but without any explanation, before his tumour was diagnosed.

During a study of quality of life already described,5 we had opportunity to visit glioma patients at home after diagnosis, to listen to their accounts, and to question relatives separately. Here we report data on 92 patients (table 1), suggesting a differing development of symptoms and problems from that described in their medical records, and a distinctly similar picture, in some, to that described by Alan Clark. Interviews tended to elicit histories of more subtle problems such as fatigue and cognitive and personality change almost as often as the neurological problems typically associated with brain tumours. Of the 48 patients with headache only two had developed no other symptoms by the time of diagnosis.

Table 1

 Symptoms at diagnosis of malignant cerebral glioma recorded in hospital records versus those elicited at home interviews

Our sample is limited to patients who were well enough for radiotherapy and to receive home visits after diagnosis. It therefore excludes those most disabled and confused at diagnosis and treated with steroids alone. The data only cover problems that had developed before diagnosis. We did not have access to primary care records to explore how symptoms were presented to general practitioners, but 41% (38 of 92) were referred to a neurologist. Of the 64 patients whom we questioned on the topic, 19% (12) were critical of the initial management by their general practitioner and 28% of 88 relatives thought there had been significant delay by the health care system as a whole. This issue remained salient for many, even after the patient had died. Of 56 whom we saw as part of a study after bereavement, one third (17) spontaneously mentioned concerns they continued to have about delay in diagnosis and the effect this might have had on quality of life or survival. The problems they identified ranged across primary, secondary, and tertiary care and included their perception that referrals had not been made quickly enough or that waiting for appointments and imaging had been excessive.

The lack of data on the development of symptoms means that current national criteria for urgent referral rely on data from patients presenting to specialised centres rather than on the predictive power of symptoms in the population attending primary and secondary care. The data elicited here confirm the earlier suggestion by McKeran and Thomas that the significance of headache may lie in its association with altered patterns of behaviour and disability.2 Although retrospective accounts cannot be used to define predictive factors for earlier diagnosis, they do suggest some implications for future research and practice. First, more detailed study of patients’ and relatives’ experience might help further define the subacute presentation of cognitive and personality change and their relation to other complaints. Second, the predictive power of neurological symptoms presenting to general practitioners could be explored using existing large primary care research datasets. Third, relatives of patients referred urgently should be asked to attend with them to clarify aspects of the history that the patient may be unaware of. Beginning to discuss openly the difficulty of earlier diagnosis may help families come to terms with this lasting aspect of their concern. This might also help repair unnecessary rifts in relations with general practitioners, who are best placed to provide local support and palliative care these patients so often need.


We thank Sue Hall and Maureen O’Connor for help collecting the data and patients and relatives for agreeing to be interviewed. Data collection was supported by the Cancer Research Campaign, grant number CP 1017, and writing up by the BMA TP Gunton Research Award for Health Education in Cancer. The views expressed do not represent those of either organisation.



  • Competing interests: none declared